What are the biggest issues with EMR today?

As medicine moves forward in its most technologically advanced era yet, we continue to struggle with basic concepts such as record keeping.  The medical record is vital to the care of the patient.  It tells the story of each patient’s journey through the medical system.  The idea of centralizing all pertinent medical information is, in theory, a step in the right direction.  In utopia, there would be one medical record for each and every patient that could be accessed by any healthcare provider on the planet at a moments notice.  However, in practice, this is a monumental task.  The New York Times published as special section on The Digital Physician.  As part of the feature, the current state of the electronic medical record (EMR) was examined.

The Federal government has mandated the implementation of EMR in order for providers to be paid at the highest allowable rates and receive certain incentive pay for complying with EMR. Terms such as “meaningful use” have been coined by legislators in Washington, DC.  Certainly, a great deal of money has been spent by both the US government as well as individual providers to develop EMR and implement electronic records by predetermined government deadlines.  EMR has the potential to provide increased patient safety and significant cost savings if developed properly.  However, current EMR systems are not really ready for “prime time”  There is no real data out there that has demonstrated improved outcomes with EMR use–it is interesting that Medicare refuses to reimburse for unproven therapies but the Federal government will mandate EMR implementation without long term outcomes data.  Only recently are studies emerging to give us some idea of the impact of EMR on patient outcomes.  I think that there is a great deal of work to be done in order for EMR to have the desired impacts on safety, communication and healthcare cost containment.

What are the biggest issues with EMR today?

1. Work flow.  Many providers find EMR to decrease efficiency and decrease patient thru-put due to non user friendly interfaces and difficult navigation.  Many physicians find that the EMR actually adds significant hours to their work day.  From personal experience, it is often that I am in the office after hours completing EMR notes.  Sometimes I have to finish them at home on the weekends.  Proponents of EMR argue that electronic records streamline and  reduce physician time spent with documentation.  Nothing could be farther from the truth.  Many EMR systems are cumbersome and time consuming.

2. Errors in documentation. Most EMRs are plagued by erroneous data.  As my wife ( a lifelong outcomes researcher) often says about databases in general–”garbage in equals garbage out”.  The point and click and drop-down menu capabilities of most EMR systems foster the propagation of erroneous data.  Unless a provider takes the time to audit the electronic record on each visit, inaccurate information can be placed in a record and continued forward.  In addition, auto-populate and recall functions can lead to documentation of things that were never done (such as particular physical exam components).

3. Interruption of the doctor-patient relationship.  As I have mentioned in a previous blog, we must take great care not to allow the computer to come between doctor and patient.  We must continue to practice the art of medicine which requires that we actually talk and listen to our patients. We must not forget the value of interacting with patients, looking them in the eye, and providing them undivided attention.  Computers, laptops, and iPads in exam rooms foster distractions–I make it a point to leave my laptop at the workstation and put my notes in the EMR after the patient has left the exam room.  This often leads to finishing notes after hours.

The future

The EMR has an important place in the practice of medicine going forward.  However, I believe that we are rushing forward, blindly running to implement flawed systems in order to meet arbitrary government deadlines.  I believe that we must take a step back, evaluate the good, the bad and the ugly about EMR.  Only then can we work to craft a system that will accomplish what everyone in medicine wants–more efficient patient care, improved outcomes, increased patient safety. 

Kevin R. Campbell is a cardiac electrophysiologist who blogs at his self-titled site, Dr. Kevin R. Campbell, MD.

Comments are moderated before they are published. Please read the comment policy.

  • http://profiles.google.com/mhirzel Mary L. Hirzel

    Why isn’t privacy #1 on your list of biggest issues? Why isn’t it on your list at all?????

    • kjindal

      the obsession with privacy and vast misinterpretation of HIPAA laws are a major contributor to poorly coordinated care between institutions, and resulting hospitalization / rehospitalization, especially among the frail elderly (ie. medicare-supported) nursing home population.
      Of course, privacy is important, but when medical office & hospital nursing station clerks refuse to transmit relevant information because “I can’t send you that because of HIPAA”, we have gone too far.
      In fact, the NUMBER ONE biggest problem with EMRs is that we don’t have a single centralized system. The medical risk and costs resulting from such a deficiency are vastly underappreciated by the programming/technical “experts”.

      • Solon Pan Zafiropoulos

        Privacy is definitely important, but what you say is true also. I have an elderly mother in this kind of situation being rehospitalized, poor coordination, etc. There should be a strong balance between BOTH privacy AND having a quality designed centralized system.

  • JPedersenB

    I agree with Mary. Why aren’t you addressing the issue of privacy? This is a big problem!

  • http://www.facebook.com/stephen.rockower Stephen Rockower

    What about the errors of “The computer made me do it that way, and I couldn’t figure out how to over-ride it”???

  • http://twitter.com/Alv_Ignacio Ignacio Alvarez

    IT always tend to change the behaviors of the new users. EMR is affecting the doctor-patient relationship because it hasn’t been thought for the users but for the management and it never considered the affects of the new system to the context. For many years now, the health system is trying to introduce the patient centered approach to the practice of medicine. EMR companies need to change the way they conceive their product by considering this approach.

    Also, I totally agree with Mary. Privacy should be the biggest issue. How can a patient thrust his doctor and the health system if he doesn’t know were his information is going. (sorry for my english)

  • CorpAvenger

    I totally agree with Mary here…. Privacy even among allowed by congress to collude insurance carriers (and therefore the gov’t since they provide a lot of healthcare insurance) among themselves to hang up doctors and patients.
    I support the exact opposite of centralization. Instead of creating centralized and all too easy to target, steal and access records, let’s create Great Inter-Operability between systems, systems of easy sharing, Modern Records transmission of very select parts, like lab results or specialists of hospital notes back and forth between providers, primary care, family doctor and the like. So providers can send, what is needed, necessary just as we presently do with snail mail and fax machines today. Not entire records allowing docs to share only what is really needed.
    By “De-Centralizing” data the increase in any given individual’s safety, the reduction of risk to be hacked and wacked goes WAY DOWN. And patients regain their privacy and sense of safety there of. This no allows for small batches to track and report, keep track of via modest sized transactions back and forth. Otherwise our personal medical data is only as safe as any clinic’s or hospital’s HR’s hiring and screening processes for all the millions of individuals who will simply by the nature of their jobs have access to these nationally interconnected records and systems. NOT a pretty thought. Actually a very scary “Big Brother” Orwellian like one to me…

  • http://www.facebook.com/harry.honeycutt1 Harry Honeycutt

    Unfortunately, this is an issue where the questions are simple, and the answers are not.

    The issues you raised (in my opinion) are almost all design issues in one guise or another, and can be resolved over a series of software releases if the proper pressure is applied to get cooperation, though it may not be in a time frame you’ll like.

    Some of the issues raised in a new system, however, may be deeper issues that the system merely exposes, troubles that were also in the previous manual system, but nobody knew that, back in the day, because the manual system didn’t show it, or because there was never anything to compare it with until now.

    You recommend that we “take a step back.” I can understand your assessment, but I disagree. Systems get better incrementally, over a series of releases, and if you “take a step back” you lose access to the information you need to make the design better.

    Like they say in the Army, no plan ever survives first contact with the enemy. If the Army metaphor doesn’t click, think about cars. Getting from the Ford model T to what you drove to the office today was not easy, and it took a lot of model years to get to. Ford did not take a step back, but he got better every year. Usually.

    You cite the arbitrary government deadlines; I would humbly suggest that they are necessary, if clumsy, because they put pressure on the people in the system to come up with something.

    I liked your insight and analysis of the problem areas for EMR, and I would address each in turn:

    1. Work flow: There is no substitute for good design. If your EMR system is difficult to work with or too time-consuming even after you are accustomed to it, then the design needs to be improved, and you need to demand that from the vendor. If he shrugs, well, maybe he’s not up to the job and you need somebody else. EMR’s are a big market: if you yell long enough, somebody will realize that you represent a market opportunity, and pick you up.

    The classic problem in “front-end” design is the choice between making something easy to use for a novice, with lots of simple choices, etc, or making something robust and powerful, and FAST (!) for a trained user who works in it every day. Unfortunately the “easy to use” option is usually the one that sells the best, since a prospective customer is generally also a new user.

    The problem is that what you really need is what you want later on, when you are closer to “power user” status. You want the more complicated (and intimidating) front end that can do more stuff faster. You might not like it, but on revision 1, you really did get what you paid for. Now, you can do more, and you want more, and you should, but you and your software vendor have to grow together, and the cost should not be all the vendor’s problem. You are learning together.

    The other classic problem in database design (it’s a medical record to you, it’s a database to me) is that a database designer is a LOT better off if she knows all the types of questions people will need to ‘ask’ of the database before she designs the database. So if you can’t find out what you need to from the records, or it takes too long, it’s possible that the database design is wrong. They can change it, but that can cost a lot of money, sometimes, and it may not be realistic unless it’s part of a release that’s a major design change for the software product.

    The other non-trivial aspect of EMR is that these records are not outside of the process, they are PART of the process, since they are the basis for billing, as well the record over which disputes are reviewed. So you are designing the process and the records for the process at the same time. The only way to do that well is incrementally, from experience, and software is just like medicine in that regard: experience ain’t cheap.

    2. Errors in documentation: this is a much more difficult problem to solve in your environment, since your only two choices are either
    1) having another human audit the records, or having a nurse enter the data and then the doctor refers to it and checks it during the visit, which introduces a cost factor that might (or might not) be a show stopper even if it was only done on a statistical sample of the records, or
    2) introducing “smart” editing rules into the front-end of the system, so that the rule could catch (hopefully) most of the input errors.
    #2 would not work reliably unless the system had a set of rules that compared fairly well to the expertise of a human doctor, which currently is only available in Star Trek episodes, though the nice people at IBM are making progress with their “Watson” projects, even the “Watson” approach may not be cost-effective for some time, either.

    Errors in logs or record keeping will also compromise the record’s usefulness in a legal setting, or other disputes, so it may actually be worth some investment to make sure the records are accurate and complete, and I would wonder whether (I don’t know) it might be highly cost-effective to be able to demonstrate that fact in, say, court.

    3. Doctor Patient relationship: My guess, this is likely one of the key problems to solve from a design standpoint, since the doctor (as we all know) needs to be able to address my medical issues as well as relate to me as a leader and teacher, and if my doctor is looking at her laptop screen, she’s not looking at me. We are people, and people need attention as well as expertise.

    I hate to give an answer that is mostly hand waving, but your EMR system needs to (as much as possible) stay out of your way and let you work the patient’s problems. Doing that well is a design challenge, to be sure, but it should be your EMR designer’s main goal, and you should press for it. With every new release, you need to see progress on this front; even though it’s hard to quantify, if it works better, you’ll know it.

    There is also the near universal problem that anyone doing a job has, from fry cooks to NFL quarterbacks, which is that (all too often) actually doing a great job and LOOKING like you are doing a great job are not exactly the same thing, so the problem with entering data while you are dealing with the patient may just be a part of this larger challenge we all face. It’s like the piano player who smiles a lot. The music isn’t any better, but the smile still helps.

    • southerndoc1

      “almost all design issues in one guise or another”
      Current EMRs are very well designed to function in a payer-centered, patient and doctor unfriendly medical system. The only way to get “better” EMRs is to fix the system. Otherwise, we’re just changing the color of the bandaid.

    • Jack Skowron

      Having 6 years experience with EMR use, using NextGen, supposedly a well regarded program, is that, while it isn’t ALL bad, it certainly does nothing to increase efficiency, good patient care, or errors. Many programs work through drop down menus…while there are complaints of handwriting issues and errors in prescribing on paper, it is so much easier to click one above or below what you wanted to get to, and, since using eRx, my errors (and pharmacy calls) have increased tremendously. I also find the hours after my day ends charting, the longer waits for the same number of patients, and the difficulty of talking and using a computer to be my experience, and EMR’s very easily can be set up to generate great looking notes, that read identically from patient to patient, and have global positives, that either have you clicking that you did things you didn’t, or drilling into the exact parts of the exam to specify, at the cost of taking significantly longer to write the note. My take on this whole EMR business is that it is not designed to help doctors or patients…it benefits administrators, insurance companies, anyone who wants to look over the doctor’s shoulder and quantify exactly how many of certain tests, prescriptions, etc, he/she is ordering. It is a bonanza for watching, and then trying to control, the practice of medicine as far as I can see, and I have been getting letters from insurance companies berating me for prescribing antibiotics for a diagnosis of “pharyngitis” (I have to enter a diagnosis on doing the test, before I have a strep result, and don’t always remember to go back and change it to “strep throat”). To me, that is really what current, “not ready for prime time” EMR use (and the pressure to adopt it), is really about.

      • southerndoc1

        Painful but well said.
        Expecting an EMR to make you more efficient or to improve patient care is like thinking you can brush your teeth with a lawn mower. That’s not what they’re designed for, and the sooner we can disabuse ourselves of this fantasy, the less frustrated we’ll be.

      • Solon Pan Zafiropoulos

        Sad but true everything you just said! It was bad enough that doctors had spent so many years schooling, that they also had to deal with the headaches of being paid fairly by insurance companies. My father was a workaholic Kidney surgeon, and I helped him from a young age on weekends and Summers. Eventually I helped him implement an early version of the Meditech EMR in the late 80′s. It was supposed to help with insurance billing, and of course migrating charts to electronic medical records for more efficiency, etc. It was a clumsy and unreliable system to say the least. Today, I tell my clients to go with PracticeFusion as it’s free, and data can be migrated out later when and if need be. Let the EMR market evolve and improve, and when it does, be it 1, 2, 5 years – spend the money on the best system that will help your practice. If you’re happy with PracticeFusion, you stay with it. ALSO, it’s a good idea to use Dragon speech in conjunction to improve efficiency.